1.204 Lecture 3 SQL: Basics, Joins SQL Structured query language (SQL) used for Data definition (DDL): tables and views (virtual tables). These are the basic operations to convert a data model to a database Data manipulation (DML): user or program can INSERT, DELETE, UPDATE or retrieve (SELECT) data. Data integrity: referential integrity and transactions. Enforces keys. Access control: security Data sharing: by concurrent users Not a complete language like Java SQL is sub-language of about 30 statements Nonprocedural language No branching or iteration Declare the desired result, and SQL provides it 1
SQL SELECT SELECT constructed of clauses to get columns and rows from one or more tables or views. Clauses must be in order: SELECT columns/attributes INTO new table FROM table or view WHERE specific rows or a join is created GROUP BY grouping conditions (columns) HAVING group-property (specific rows) ORDER BY ordering criterion ASC DESC Example tables Orders Customers OrderNbr Cust Prod Qty Amt Disc 1 211 Bulldozer 7 $31,000.00 0.2 2 522 Riveter 2 $4,000.00 0.3 3 522 Crane 1 $500,000.00 0.4 CustNbr Company CustRep CreditLimit 211 Connor Co 89 $50,000.00 522 AmaratungaEnterprise 89 $40,000.00 890 Feni Fabricators 53 $1,000,000.00 RepNbr Name RepOffice Quota Sales SalesReps 53 Bill Smith 1 $100,000.00 000 $0.0000 Offices 89 Jen Jones 2 $50,000.00 $130,000.00 OfficeNbr City State Region Target Sales Phone 1 Denver CO West $3,000,000.00 $130,000.00 970.586.3341 2 New York NY East $200,000.00 $300,000.00 212.942.5574 57 Dallas TX West $0.00 $0.00 214.781.5342 2
Example schema +OrderID Using SQL Server and Management Studio Express Your SQ L Server database engine should start by default when your system starts Ask TA for help if needed at office hours Start Management Studio Express (MSE) from Start- >Programs->SQL Server 2008 Open Lecture3CreateDB.sql with MSE in Windows Explorer Download the.sql file from Stellar and double-click on it Select Execute from toolbar Database MIT1204 will be created and data inserted for examples during this class Open Lecture3Examples.sql for all SQL code in this lecture Experiment with it 3
SQL queries: SELECT Click New Query in MSE; type these statements: List the sales reps SELECT Name, Sales, Quota FROM SalesReps Find the amount each rep is over or under quota SELECT Name, Sales, Quota, (Sales-Quota) FROM SalesReps Find the slackers SELECT Name, Sales, Quota, (Sales-Quota) FROM SalesReps WHERE Sales < Quota RepNbr Name RepOffice Quota Sales 53 Bill Smith 1 $100,000.00 $0.00 89 Jen Jones 2 $50,000.00 $130,000.00 SQL queries: calculation, insert, delete, update Find the average sale SELECT AVG( G(Amt) FROM Orders; Find the average sale for a customer SELECT AVG(Amt) FROM Orders WHERE Cust = 211; Add an office INSERT INTO Offices (OfficeNbr, City, State, Region, Target, Sales, Phone) VALUES ( 55, Dallas, TX, West, 200000, 0, 214.333.2222 ); Delete a customer DELETE FROM Customers WHERE Company = Connor Co ; (Syntax is valid but command will fail due to referential integrity) Raise a credit limit UPDATE Customers SET CreditLimit = 75000 WHERE Company = Amaratunga Enterprises ; 4
SELECT: * and duplicates Select all columns (fields) SELECT * FROM Offices; Duplicate rows: query will get two instances of West SELECT Region FROM Offices; Eliminate duplicates: SELECT DISTINCT Region FROM Offices; NULLs NULL values evaluate to NOT TRUE in all cases. Insert NewRep with NULL (blank or empty) Quota The following two queries will not give all sales reps: SELECT Name FROM SalesReps WHERE Sales > Quota; SELECT Name FROM SalesReps WHERE Sales <= Quota; A new rep with a NULL quota will not appear in either list Check for NULLS by: SELECT Name FROM SalesReps WHERE Quota IS NULL; 5
SELECT Operators SELECT * FROM <table> WHERE Disc*Amt > 50000; (Orders) WHERE Quota BETWEEN 50000 AND 100000; (SalesReps) Range is inclusive (>=50000 and <=100000) WHERE State IN ( CO, UT, TX ); (Offices) WHERE RepNbr IS NOT NULL; (SalesReps) WHERE Phone NOT LIKE 21% ; (Offices) SQL standard only has 2 wildcards % any string of zero or more characters (* in Access) _ any single character (? in Access) Most databases have additional/different wildcards. SQL Server has: [list] [^list] match any single character in list, e.g., [a-f] match any single character not in list, e.g. [^h-m] SELECT: COUNT, GROUP BY Parts PartID Vendor 123 A 234 A 345 B 362 A 2345 C 3464 A 4533 C Number of parts from vendor A SELECT COUNT(*) FROM Parts WHERE Vendor = A ; Result: 4 Number of parts from each vendor SELECT Vendor, COUNT(*) AS PartsCount FROM Parts GROUP BY Vendor; Result: Vendor PartsCount A 4 B 1 C 2 6
Examples What is the average credit limit of customers whose credit limit is less than $1,000,000? SELECT AVG(CreditLimit) FROM Customers WHERE CreditLimit < 1000000; How many sales offices are in the West region? SELECT Count(*) FROM Offices WHERE Region= 'West ; Increase the price of bulldozers by 30% in all orders UPDATE Orders SET Amt= Amt*1.3 WHERE Prod= 'Bulldozer ; Delete any sales rep with a NULL quota DELETE FROM SalesReps WHERE Quota IS NULL; Joins Relational model permits you to bring data from separate tables into new and unanticipated relationships. Relationships become explicit when data is manipulated: when you query the database, not when you create it. This is critical; it allows extensibility in databases. You can join on any columns in tables, as long as data types match and the operation makes sense. They don t need to be keys, though they usually are. Good joins Join columns must have compatible data types Join column is usually key column: Either primary key or foreign key Nulls will never join 7
Joins List all orders, showing order number and amount, and name and credit limit of customer Orders has order number and amount, but no customer names or credit limits Customers has customer names and credit limit, but no order info SELECT OrderNbr, rnbr, Amt, Company, CreditLimit FROM Customers, Orders WHERE Cust = CustNbr; (Implicit syntax) SELECT OrderNbr, Amt, Company, CreditLimit FROM Customers INNER JOIN Orders ON Customers.CustNbr = Orders.Cust; (SQL-92) OrderNbr Cust Prod Qty Amt Disc 1 211 Bulldozer 7 $31,000.00 0.2 2 522 Riveter 2 $4,000.00 03 0.3 3 522 Crane 1 $500,000.00 0.4 Join CustNbr Company CustRep CreditLimit 211 Connor Co 89 $50,000.00 522 Amaratunga Enterprises 89 $40,000.00 890 Feni Fabricators 53 $1,000,000.00 Join with 3 tables List orders over $25,000, including the name of the salesperson who took the order and the name of the customer who placed it. SELECT OrderNbr, Amt, Company, Name FROM Orders, Customers, SalesReps WHERE Cust = CustNbr AND CustRep = RepNbr AND Amt >= 25000; (Implicit syntax) OrderNbr Cust Prod Qty Amt Disc 1 211 Bulldozer 7 $31,000.00 0.2 2 522 Riveter 2 $4,000.00 0.3 3 522 Crane 1 $500,000.00 0.4 Join Join CustNbr Company CustRep CreditLimit 211 Connor Co 89 $50,000.00 522 890 Amaratunga Enterprises Feni Fabricators 89 53 $40,000.00 $1,000,000.00 RepNbr Name RepOffice Quota Sales 53 Bill Smith 1 $100,000.00 $0.00 89 Jen Jones 2 $50,000.00 $130,000.00 Result: OrderNbr Amt Company Name 1 $31,000.00 Connor Co Jen Jones 3 $500,000.00 AmaratungaEnterprise Jen Jones 8
Join notes SQL-92 syntax for previous example: SELECT OrderNbr, Amt, Company, Name FROM SalesReps INNER JOIN Customers ON SalesReps.RepNbr = Customers.CustRep C t INNER JOIN Orders ON Customers.CustNbr = Orders.Cust WHERE Amt >= 25000; Use * carefully in joins It gives all columns from all tables being joined If a field has the same name in the tables being joined, qualify the field name: Use table1.fieldname, table2.fieldname Customers.CustNbr, Orders.Amt, etc. You can join a table to itself (self-join). See text. JOIN types INNER join: returns just rows with matching keys (join column values) RIGHT join: returns all rows from right (second) table, whether they match a row in the first table or not LEFT join: returns all rows from left (first) table, whether they match a row in the second table or not OUTER join: Returns all rows from both tables, whether they match or not 9
Examples List customer names whose credit limit is greater than their sales rep s quota. Also list the credit limit and quota. SELECT CreditLimit, Quota, Company FROM SalesReps INNER JOIN Customers ON SalesReps.RepNbr = Customers.CustRep WHERE CreditLimit>Quota; List each rep s name and phone number SELECT Name, Phone FROM Offices INNER JOIN SalesReps ON Offices.OfficeNbr = SalesReps.RepOffice; Display all customers with orders or credit limits > $50,000. SELECT DISTINCT CustNbr FROM Customers LEFT JOIN Orders ON CustNbr = Cust WHERE (CreditLimit > 50000 OR Amt > 50000) 10
MIT OpenCourseWare http://ocw.mit.edu 1.204 Computer Algorithms in Systems Engineering Spring 2010 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.